This post summarises the findings of a paper by Barbara Tversky: Gestures for Thinking and Explaining (2005). The author demonstrates how we use gesture, even when we are not communicating, to solve problems, especially when those questions put load onto our working memories.

The paper makes use of two problems: the two string problem and the six cup problem.

The two string problem was developed to explore spatial problem solving (Maier, 1931). The participant is asked to tie the ends of the two strings together. However, if she holds one end, she is unable to reach the other string. Even if the problem is challenging, the load it places on working memory is low: there are few factors to remember.

In comparison, the six cup problem (Ashcraft, 1994), puts a larger load on working memory. The solver needs to remember the position of six cups and whether they contain water to hold in memory.

The challenge is to move one cup only so that alternate cups are full of water.

22 undergraduates were videoed solving the two problems silently. The number of gestures they made while solving the problem was recorded.

The chart shows that when the working memory demand increases, the participants make use of far more gesture. Note – this isn’t about communication – the participants were being videoed, but they were not engaged with the camera. The gestures appear to support thinking.

Once the participants had solved each problem, they were asked to present the solution to the camera.

The number of enactment gestures – physically demonstrating the actions in solving the problem – were high whether the working memory load was high or low. The number of scene creation gestures increased significantly with the complexity of the problem.

What does this tell us?

Tversky’s beautiful 8th law of cognition states: When thought overflows the mind, the mind puts it back into the world.

This experiment shows thought overflowing the mind into the world through gesture. Even when not communicating, we use gesture to help us solve problems, especially when there is high load on working memory.

We also rely on gesture to communicate a solution.

So a question: if we use gesture to solve problems and communicate a solution, should we recognise and encourage gesture in class in the same way we would encourage the use of working out and diagrams?